|Número de publicación||US7283253 B2|
|Tipo de publicación||Concesión|
|Número de solicitud||US 10/389,269|
|Fecha de publicación||16 Oct 2007|
|Fecha de presentación||13 Mar 2003|
|Fecha de prioridad||13 Mar 2002|
|También publicado como||EP1606609A1, US20030222197, US20060039010, US20080245953, WO2004083832A1|
|Número de publicación||10389269, 389269, US 7283253 B2, US 7283253B2, US-B2-7283253, US7283253 B2, US7283253B2|
|Inventores||Steven A. REESE, Carl S. BROWN, Paul C. GOODWIN|
|Cesionario original||Applied Precision, Llc|
|Exportar cita||BiBTeX, EndNote, RefMan|
|Citas de patentes (11), Otras citas (2), Clasificaciones (14), Eventos legales (7)|
|Enlaces externos: USPTO, Cesión de USPTO, Espacenet|
The present application is a continuation in part and claims the benefit of priority of non-provisional Application No. 10/215,265, filed Aug. 6, 2002, entitled “TIME-DELAY INTEGRATION IMAGING OF BIOLOGICAL SPECIMENS,” and also claims the benefit of priority of co-pending U.S. provisional application Ser. No. 60/364,762, filed Mar. 13, 2002, entitled “METHOD FOR MULTI-AXIS INTEGRATION (MAI) IMAGING OF THICK SPECIMENS,” and claims the benefit of priority of co-pending U.S. provisional application Ser. No. 60/431,692, filed Dec. 6, 2002, entitled “OPTICAL AXIS INTEGRATION SYSTEM AND METHOD, which three applications are incorporated by reference.
Aspects of the present invention relate generally to image acquisition systems and image processing techniques, and more particularly to a system and method of integrating image data obtained along an optical axis and in conjunction with lateral translation of an image plane.
A variety of autofocus techniques have been developed to enable automated and semi-automated scanning of three-dimensional (3D) objects using short depth-of-field (DOF) optics. Such short DOF optical systems are routinely used in microscopes and other inspection systems, at least partially because a short DOF can provide superior image quality and resolution within the limits of the DOF. By way of background, the resolution of an image is directly proportional to the numerical aperture (NA) of the optics; consequently, high resolution images generally require high NA values. As indicated in Table 1, however, DOF is inversely proportional to the NA; accordingly, a high NA (i.e., high resolution) necessarily generates a narrow DOF which, in turn, increases blurring in the final image since portions of the imaged object extend above and below the plane of focus, or image plane.
In that regard, out-of-focus images are usually not useful for image analysis and are typically a cause of failure in automated scanning and image acquisition systems. Accordingly, autofocus features have typically been regarded as essential components of many conventional, short DOF automated systems.
Autofocus is not a trivial technology, however, and as a consequence, incorporating autofocus functionality presents a variety of complicated challenges associated with design, implementation, integration, and use in conjunction with common imaging systems. In particular, automatically determining a desired or optimal focal plane can be an impossible task when an object to be viewed or imaged is larger in the axial dimension than the DOF of the optics (see Table 1). In such cases, the object to be imaged generally has a plurality of focal planes, none of which may be “correct” or optimal. Even when an autofocus feature is successful in ascertaining a “best” focal plane, the time required to determine that plane of best focus can be a major limitation with respect to the scan rate.
DOF examples (at 540 nm wavelength) for numerical
apertures (NA) from 0.10 to 1.40.
Low NA air
Med NA air
Med NA air
Med NA air
High NA air
High NA water
High NA oil
High NA oil
As is generally known in the art, autofocusing methods may be employed in several ways. For example, autofocus techniques are typically used to collect images of two-dimensional (2D) objects mounted or disposed on non-planar surfaces. Further, autofocus is also commonly used to collect images of 3D objects; in fact, most objects of interest with respect to operation of typical image acquisition systems are 3D rather than simply 2D. In some situations, any 2D image of the object may be considered sufficient. In other situations, however, an image captured from a particular, predetermined, or otherwise specified focal plane is desired. In either case, the 2D image obtained with an autofocus technique is ordinarily expected to contain valuable morphological and/or quantitative information about the object, which is usually 3D, as noted above.
Additionally, autofocus may be employed to locate a starting point for a series of images to be obtained along the optical axis (so called “optical sections”); using image data from sequential optical sections, an image processing system may ascertain more information from the resulting 3D image than is possible from analysis of the individual corresponding 2D images. It will be appreciated, however, that each optical section generally contains information from neighboring sections (i.e., sections above and below the focal plane of a given section) due to the DOF range for a given NA, as illustrated in Table 1.
Images obtained from an autofocused system are typically analyzed with algorithms that are based on some form of intensity integration. For example, the pixel intensities within a particular region of interest (e.g., on a microscope slide or a microarray) can be summed or totaled to infer the quantity of a particular chemical within that region; additionally or alternatively, pixels can be counted to determine the total area of the region. At the most basic level, certain existing tests are designed to identify the simple presence or absence of a particular chemical; these tests generally rely upon an intensity integration and are configured to yield a binary result.
In any event, conventional systems generally rely upon the sequential acquisition of a plurality of optical sections followed by a computationally expensive deconvolution operation. In particular, the deconvolution is unnecessarily inefficient, since each particular optical section contains blurred image data from neighboring sections as noted above.
Embodiments of the present invention overcome the above-mentioned and various other shortcomings of conventional technology, providing an image acquisition system and method employing integration of image data obtained along an optical axis in conjunction with synchronized translation in a lateral direction. A system and method operative in accordance with some embodiments, for example, may integrate (or project) image data in the z direction as the data are acquired; significant savings in computational overhead may be achieved by projecting the image data prior to deconvolution. Lateral translation of the image plane during the scan in the z direction may provide additional efficiency, allowing large areas to be imaged in a single scan sequence.
Generally, an image acquisition system and method employing multi-axis integration (MAI) in accordance with the present disclosure may incorporate both optical axis integration (OAI) and time-delay integration (TDI) techniques.
In accordance with some exemplary embodiments, a method of acquiring data comprises: scanning an object along an optical axis; simultaneously scanning the object along a lateral axis; acquiring image data of the object during the scanning and the simultaneously scanning; and integrating the image data concomitantly with the acquiring. The scanning may comprise providing relative translation along the optical axis of the object and an image plane; similarly, the simultaneously scanning may comprise providing relative translation along the lateral axis of the object and an image plane.
Systems and methods are disclosed wherein the acquiring comprises utilizing a charge-coupled device. In some methods, the integrating comprises utilizing an image processor, and may further comprise deblurring the image data subsequent to the integrating, deconvolving the image data subsequent to the integrating, or both.
In accordance with some exemplary implementations, a method of acquiring data may further comprise calculating a two-dimensional projection of the object from a projection of the image data and a projection of an optical point-spread-function.
The scanning may further comprise selectively alternating a direction of the relative translation along the optical axis. Additionally, the simultaneously scanning may comprise synchronizing the relative translation along the lateral axis with a rate associated with the acquiring.
As set forth in more detail below, a method of acquiring image data of an object may comprise: performing an optical axis integration scan; simultaneously executing a time-delay integration scan sequence; and selectively repeating the performing and the executing.
In some embodiments, the performing comprises acquiring image data of the object at an image plane positioned along an optical axis; the performing may further comprise providing relative translation along the optical axis of the object and the image plane; as noted briefly above, a direction of the relative translation may be selectively alternated.
The executing generally comprises providing relative translation along a lateral axis of the object and the image plane; in some implementations, the executing comprises synchronizing the relative translation along the lateral axis with a data acquisition rate associated with an imaging device.
As set forth above, systems and methods are disclosed wherein: the acquiring comprises utilizing a charge-coupled device; the performing further comprises integrating the image data concomitantly with the acquiring; the integrating comprises utilizing an image processor; or some combination thereof.
Image acquisition systems and methods in accordance with the present disclosure may comprise deblurring the image data subsequent to the integrating, deconvolving the image data subsequent to the integrating, or both. A method may further comprise calculating a two-dimensional projection of the object from a projection of the image data and a projection of an optical point-spread-function.
The foregoing and other aspects of various embodiments of the present invention will be apparent through examination of the following detailed description thereof in conjunction with the accompanying drawings.
As noted briefly above, image acquisition throughput often represents the rate-limiting factor in systems and methods of scanning high-content and high-throughput assays common in biomedical and other applications. Image acquisition throughput can be especially problematic when an assay requires detection of fluorescent probes, for example, and when high lateral resolution (in the x and y dimensions) is required for high-content image analysis algorithms. In cases where the detected signal is weak such as in fluorescence imaging for example, high numerical aperture (NA) lenses are generally used to maximize collection efficiency and to minimize exposure time. A side effect of high NA lenses, however, is that the depth-of-field (DOF, or the dimension of the in-focus region measured in the z direction) is very shallow. As set forth above, high NA lenses have limited ability to view thick objects, and are unable to follow uneven substrates without refocus.
Even in cases where the detected signal is strong or is otherwise easily acquired (such as transmitted visible light, for example) optical systems can still perform inadequately if the sample thickness is greater than can be imaged by the optical DOF; additional imaging difficulties can be introduced if the object to be imaged is not located in a plane orthogonal to the optical axis. These optical limitations often lead to the use of autofocus technology, or the need to acquire images at more than one focal plane.
Although much effort has been invested in autofocus technologies, optical axis integration techniques are more cost effective and generally provide improved performance in many scanning applications. The scanning techniques set forth in detail below are very tolerant of objects having inconsistent or variable focal planes, for example, and may be used to image thick objects. Additionally, scans performed in accordance with the present disclosure may be faster than those implementing autofocus or optical sectioning procedures.
Optical Axis Integration
As an alternative to conventional autofocus methodologies, a system and method operative in accordance with the present disclosure employ optical axis integration (OAI) techniques as set forth in detail below. For a particular object to be imaged, for instance, rather than attempting to determine a particular focal plane for optics or an imaging apparatus (i.e., precisely determining an appropriate or optimal z position of the image plane), the object may be scanned along the optical axis while a detector, computer, or other computational apparatus concomitantly integrates the acquired images or image data. The resulting image is an integral (i.e., projection) of the image of the three-dimensional (3D) object along the optical axis. That is, an OAI image may generally be expressed as follows:
where i′ is the two-dimensional (2D) projection of a 3D image, i, along the optical axis (z direction).
In this context, the 3D image, i, can be described mathematically as the object (o) of interest convolved with the point-spread-function (PSF) of a microscope or other optical apparatus, as follows:
Inserting equation 2 into equation 1 gives
Rearranging the integration along the optical axis, z, then yields
which is equivalent to
Rearranging the integration along z′, the OAI image, i′(x,y), may be expressed as:
Equation 6 shows that an OAI image, i′(x,y), may be expressed as the convolution of the integral of the object along the optical axis with the integral of the PSF along the optical axis. Equation 6 is also illustrative of the relationship between the projection of the object, the projection of the image, and the projection of the PSF along the optical axis.
The following definitions may facilitate further simplification of the foregoing formulation:
Inserting the definitions expressed above into Equation 6 yields
The best method of solving Equation 8 for o′(x,y) involves Fourier Transforms, and is a well known procedure. Applying a Fourier Transform to both sides of Equation 8 and applying the convolution theorem (see, e.g., Bracewell, 1986) results in the following relationship:
Capital letters have been used to denote the Fourier Transform of the corresponding functions, and the Fourier Transform of the PSF has been replaced with the conventional term for its Transform, the optical transfer function (OTF). Rearranging terms and performing an inverse Fourier Transform then gives
where F−1 represents the inverse Fourier Transform.
Equation 10 describes an efficient method of calculating a 2D projection of an object from a projection of the image and a projection of the optical PSF. A single-step solution may work well with good quality images; for lower quality images, however, an iterative solution of Equation 10 may yield a more reliable result. See, e.g., the constrained iterative technique described by Agard et al. (David A. Agard and John W. Sedat, Nature, volume 302, 1984, pages 676 et seq.).
As described and contemplated in the present disclosure, a system and method may implement, incorporate, or comprise OAI techniques in either of two forms: digital; or analog. In embodiments incorporating or practicing digital OAI, for example, a series of images may be collected along the optical axis and then digitally summed to form i′(x,y). This summation may occur during or after the scan, i.e., it may not be necessary to save individual optical sections as discrete images or collections of image data. In analog OAI embodiments, for example, i′(x,y) may be generated by scanning the object along the optical axis while the image data are accumulated within a charge-coupled device (CCD) or other detector. The integration may be performed in the CCD chip and generally may result in only a single image, i.e., a single image may represent the entire depth of the object in the z direction along the optical axis.
Analog OAI may have particular utility with respect to operations involving scanning microarrays, for example, with CCD cameras or other detectors. A system and method employing analog OAI may eliminate or substantially reduce reliance upon sophisticated, time-consuming, and processor intensive autofocus procedures.
In many applications, analog OAI may provide a number of advantages over digital OAI and autofocus, especially for automated scanners. For example, as compared with digital OAI, the analog OAI embodiments: require substantially lower data collection and processor overhead; exhibit lower read noise; and exhibit lower photon noise for equivalent exposure times.
As compared with traditional autofocus systems, advantages of the analog OAI embodiments may include the following: faster scan times; lower total exposure requirements; minimization or elimination of problems related to determining an arbitrary plane of focus; and integration of the 3D object yields or allows full quantitation of the object, i.e., the information content of the OAI image is higher than that achieved with autofocus systems, and accordingly, fewer structures associated with the object of interest are missed.
As compared with the analog technology, advantages of digital OAI embodiments may include a potential for achieving substantially larger photon counts; accordingly, 3D images may be made available for advanced image analysis such as 3D deconvolution, volumetric measurements, and the like.
The synergistic combination of the OAI techniques described above with deconvolution, for example, may provide a significant advance for automated slide scanning techniques. For instance, OAI images generally may benefit from the quantitative deblurring procedure; similarly, deconvolution performance may be improved because Equation 10 deals with images in 2D rather than 3D. Furthermore, many forms of image analyses based upon images obtained from autofocused systems will work equally well (or better) with projected images.
For example, a basic object detection operation may benefit from OAI image processing techniques; in that regard, it will be appreciated that images with minimal DOF (i.e., autofocus images) are less likely to contain a specific object of interest than the corresponding projection image. Likewise, analyses that use intensity integration may also benefit from application of OAI techniques, at least because the z dimension (i.e., along the optical axis) is already integrated into the OAI result. By way of another example, assays that integrate intensities within 3D structures (e.g., nucleus, cytoplasm, and endoplastic reticulum) may generally be more accurate with OAI images because 2D autofocus images cannot properly measure out-of-focus intensities.
Turning now to the drawing figures,
System 100 generally comprises a microscope operably coupled to a precision movable stage 120 and to an image acquisition component 140; stage 120 may be configured and operative to support a microarray, microscope slide, or other similar structure (reference numeral 190) upon which a specimen or object 199 to be imaged is disposed. As is generally known in the art, microscope 110 may comprise, or be operative in conjunction with, an illumination source 111 for illuminating stage 120, slide 190, or both with light of a predetermined or selected frequency or spectral bandwidth; in that regard, illumination source 111 may provide light in the visible, infrared, or ultraviolet wavelengths.
In some embodiments, illumination source 111 may be incorporated within housing 112 of microscope 110, i.e., on the opposite side of stage 120 and slide 190 than depicted in
As noted above, stage 120 may be movable relative to optics (e.g., objective 119 illustrated in
In some embodiments, stage 120 may also be movable along the z axis (the optical axis). It will be appreciated that microscope optics may also facilitate positioning an object on slide 190 in the proper location in 3D space (i.e., x, y, and z coordinates) relative to the optical path and the focal point of objective 119. In that regard, one or more optical components of microscope 110 such as objective 119 may be movable in the z direction, either in addition to, or as an alternative to, selectively moving stage 120 along the optical axis. Additionally or alternatively, objective 119 may be movable along the x axis, the y axis, or both.
It will be appreciated that numerous mechanisms and methods of positioning object 199 to be imaged relative to microscope optics are generally known. Relative movement of various components (such as slide 190, stage 120, and objective 119, for example), either individually or in combination, may vary in accordance with system requirements and configuration, and may be effectuated to position object 199 in a suitable location relative to objective 119. The present disclosure is not intended to be limited by the structures and processes employed to position object 199 relative to objective 119 and the optical path or the image plane. Accordingly, reference made herein to relative motion of object 199 and an image plane may generally comprise movement of object 199, movement of the image plane, or some combination of both.
Microscope optics may generally be configured and operative in conjunction with image acquisition component 140; in that regard, component 140 generally comprises a camera, charge-coupled device (CCD), or other detector 141 operably coupled to an image processor 142 or other appropriate electronics. System 100 may additionally include control electronics 150 operative to control, for example: operational parameters, functional characteristics, or other configurable aspects of image processor 142 and detector 141; two- or three-dimensional motion of stage 120, objective 119, or other components; power output, spectral bandwidth, frequencies, or other parameters for source 111 and any other illumination source incorporated into system 100; data storage; and the like. In that regard, electronics 150 may comprise one or more microprocessors, microcontrollers, or other programmable devices capable of executing computer readable instructions; additionally, electronics 150 may also comprise or be operably coupled with data storage media or networked devices such as file servers, application servers, and the like. Those of skill in the art will appreciate that various methods and apparatus employing microprocessors or computer executable instruction sets to configure and to control operation of image acquisition systems are generally known.
In operation, image data acquired by detector 141 may be summed, manipulated, saved, or otherwise processed by hardware, software, or both resident at image processor 142; in some embodiments, functionality of processor 142 may be influenced or controlled by signals transmitted from electronics 150 as noted above. Alternatively, the functionality of image processor 142 and electronics 150 may be incorporated into a single device, for example. Specifically, image processor 142 may be operative in accordance with instruction sets to compute solutions or approximations for the equations set forth herein.
In both the analog (
In the analog embodiment of
As noted above, an analog method of OAI may comprise obtaining image data (block 414) continuously during the z axis scan (i.e., relative translation of the object and the image plane along the optical axis). Specifically, the OAI image may be generated by scanning the object along the optical axis while the image data are accumulated within a CCD or other detector 141. Accordingly, the integration or projection may be performed during (concomitantly or substantially simultaneously with) data acquisition, and the image data written to memory at block 416 generally may represent a single image, i′(x,y), such as depicted in the center of
In contrast to the analog method, the digital OAI embodiment of
In accordance with a determination made at decision block 431, the method may progress to the next optical section in sequence, moving the object, the microscope optics, or both, so as to position the focal plane at the next sequential position along the z axis of the object; this event is represented at block 427. Following a progression through a desired, selected, or predetermined range in the z direction (as determined at decision block 431), the method may end (block 428), resulting in or allowing generation of a single, two-dimensional OAI image, i′(x,y), representing the entire depth of the object in the z direction. As noted above, the method of
As used herein, the phrase “time-delay integration” (TDI) generally represents a method of continuous scanning which may be implemented in conjunction with CCD cameras or other imaging devices. In CCD cameras, for example, incident light is creates electric charge at individual charge-coupled wells on the device surface. Charged electrons are then transferred sequentially down the columns of the chip (parallel shifts) while the row that reaches the bottom of the chip is transferred to an accumulator called the serial register. The serial register is then shifted horizontally and processed by an A/D converter.
In accordance with some TDI embodiments, precision motion control may be employed to synchronize motion of the object being imaged or motion of the camera or other imaging device (as set forth above with reference to
It will be appreciated that the
In this context, synchronous motion between the object and the CCD row may be effectuated substantially as set forth in detail above with reference to
Upon examination of
Various embodiments of TDI may be employed to image objects in applications involving limited or minimal signal intensity. Specifically, the motion control characteristics of TDI allow for longer exposure per picture element (pixel) for a given total image collection time, facilitating increased imaging quality even in instances where signal intensity ordinarily would be a deleterious factor. In many biological sample imaging applications, for example, signal intensity may be limited or otherwise impeded either by the nature of the sample, by the illumination intensity, by the physical characteristics or operational parameters of the indicators used in conjunction with the sample, or by some combination of the foregoing.
By way of example, one application in which the above-mentioned factors are especially problematic is in the imaging of fluorescently labeled biological specimens. In imaging applications involving such samples, all three limitations noted above (ie., related to the sample, the illumination source, and the indicator employed) are prevalent. Accordingly, TDI methodologies may be used in conjunction with known fluorescence imaging technology to minimize the attendant effects of weak signal intensities.
While it will be appreciated that TDI techniques may prove useful in the context of multiple panel collection imaging schemes producing images such as those illustrated in
As mentioned above, one of the difficulties associated with scanning biological specimens, especially for fluorescence characteristics, is generally due to the fact that a finite limit exists with respect to the intensity of light emanating from an illuminated sample. In accordance with the present disclosure, however, adjusting the scan rate (i.e., the relative movement of the sample across the CCD imaging surface) and the readout speed of the imaging device enables a system and method of TDI imaging to control the exposure time for each pixel in the acquired image.
Additionally, it will be appreciated that many samples (in biological fields and in other scientific areas) are labeled with multiple indicators, each of which may be spectrally separated the others. Consequently, some TDI embodiments may incorporate an ability to address multiple wavelength data. By way of example, data spanning multiple wavelengths may be acquired in at least two different ways: sequential scanning; and simultaneous scanning.
In TDI implementations incorporating sequential scanning techniques, a single, monochromatic detector may be employed; in this embodiment, multiple wavelengths may be selected through filters, for instance, or through other wavelength selection components or methods. To construct a single, multiple wavelength image, an instrument or system operative in accordance with the present disclosure may scan the sample (or a strip of the sample) through a selected or predetermined filter to acquire image data at a desired wavelength, change the filter (using a filter wheel, for example, or similar apparatus), and scan the same sample (or strip of sample) through the new filter to acquire additional image data at a different desired wavelength. The foregoing operations may be selectively repeated in accordance with system requirements to acquire a plurality of scan images at a desired number of selected wavelengths. In this manner, a final image may be constructed of the data acquired during the plurality of sequential scans.
It will be appreciated that one of the challenges associated with such a methodology is the registration of scans, particularly in instances where the exposure time used for one wavelength differs from the exposure time used for another. In such situations, a TDI system and method may measure and control the actual velocities and positions of the sample during each of the plurality of scans; precise control of scan speeds, and accurate measurements thereof, may prevent or allow for correction of chromatic shift in the portions of the image that are derived from the component scans. Accordingly, systems and methods of TDI as described herein may selectively synchronize the movements as set forth above responsive to such different conditions.
In some embodiments, multi-spectral image data may be acquired from a single sample using multiple detectors, for example, each with its own spectral response. In this manner, a multiple wavelength image may be collected in a single scan. Alternatively, such a multi-spectral scan may be accomplished with a single detector or imaging device equipped with a specially designed color mask. In that regard,
It is noted that positioning methodologies in various TDI embodiments may employ constant relative velocity of the object to be imaged and the imaging surface of the imaging device. As set forth above, constant relative positioning of the object to be imaged may be accomplished through precise motion of the slide, the stage, the optics, or some combination thereof. In addition, it is possible to implement TDI methods employing one or more incremental positioning strategies; in that regard, relative motion of the object to be imaged may be synchronized to the readout (or output) capabilities of the CCD camera or other imaging device. In this manner, long exposures may be accommodated.
In such embodiments, the velocity of the object may become so slow as to become non-constant, depending upon the output bandwidth and readout rate of the imaging device. In some circumstances, for example, the sample or object may be translated (relatively) a distance corresponding to one camera row, maintained at that location for the duration of an exposure, and subsequently translated a distance corresponding to the next camera row. The foregoing procedures may be selectively repeated in accordance with system requirements or as desired. As set forth above, relative movement of the object, the slide, the stage, the optics, or some combination thereof, may begin and cease at each row shift in the camera or imaging device. The high degree of synchronization provided by such an approach may yield excellent resolution and registration between wavelengths.
Rotational and positional errors in precision motion stage systems may be measured and corrected, for example, using optically determined coordinate calibration or other methods. In particular, such errors may be minimized dynamically through precise modification of stage movements. Specifically, such coordinate calibration may also be used in conjunction with. TDI techniques dynamically to correct rotational and positioning errors during a given TDI scan.
As noted generally above, a system and method employing multi-axis integration (MAI) may incorporate both OAI and TDI techniques.
In that regard, TDI is described in co-pending U.S. nonprovisional application Ser. No. 10/215,265, filed Aug. 6, 2002, entitled “TIME-DELAY INTEGRATION IMAGING OF BIOLOGICAL SPECIMENS.” Specifically, TDI methodologies as set forth in detail above enable a system and method of image data processing simultaneously to collect image data over a given lateral area and to read data out of the camera or detector. That is, data are acquired and read from the detector simultaneously. Accordingly, a system implementing TDI methods may efficiently acquire image data over a large area (relative to the focal plane of the detector) with substantially reduced overhead as compared to systems which do not implement TDI.
U.S. provisional application Ser. No. 60/431,692, filed Dec. 6, 2002, entitled “OPTICAL AXIS INTEGRATION SYSTEM AND METHOD” describes various methodologies for collecting image data as an integration (along the optical axis) of intensity information for a given x-y image frame. As set forth in detail above with reference to
While the OAI methodologies set forth above are powerful, the disclosed embodiments may be augmented with one or more TDI scanning techniques. As set forth above with reference to
In that regard, a system and method incorporating MAI techniques may generally acquire a plurality of OAI images, each of which may include data captured from a different lateral location in x-y space on slide 190 or stage 120. The lateral scan may employ a raster or serpentine pattern, for example; in some embodiments, a serpentine lateral scan pattern may enable an MAI system and method to operate at optimum efficiency.
Various data collection parameters may be optimized to maintain quantitative accuracy in accordance with system requirements. In particular, z-axis motion may be suitably controlled to assure an appropriate scan profile; accordingly, the resulting OAI image may not be biased with information from any particular z location in the sample or the object to be imaged. Additionally, the period of the scan profile may be computed as a multiple of the time required to read a whole frame of data from detector 141 as set forth above in detail with reference to
Various MAI techniques may incorporate either continuous (analog) or incremental (digital) scanning methodologies, as set forth below.
In accordance with some embodiments employing an analog MAI technique, the z location of the image plane may be moved continuously along the optical axis between a minimum and maximum position (i.e., in and out of focus). The velocity of this z translation may be determined or calculated as a function of the required scan distance (i.e., the distance between the minimum and maximum z locations of the image plane), as well as a desired scan time period. Acceleration may be made as large as possible such that, for example, the position-versus-time curve of the image plane may have a triangular wave pattern. A system operating in accordance with the analog MAI scan pattern illustrated in
During translation of the image plane through the z scan, the y axis may be simultaneously scanned at constant velocity such that the image traverses the entire detection surface of detector 141 while the z translation completes a cycle. Accordingly, each row of detector 141 may be exposed to each z position four times as indicated by the points marked za in
In accordance with digital MAI methodologies, the y and z axis translations may be executed incrementally while rows of image data are read from the detector.
By way of example and not by way of limitation, the number of increments, nz, along the z axis in some embodiments of a digital MAI cycle may be expressed as
n z=3n y−1
where ny represents the number of lateral MAI increments in a single MAI cycle. In the case of a CCD detector, for example, ny may simply represent the number of binned CCD rows:
In some implementations, the scan may also be constrained by the following equation for nz:
n z=(total scan range along z)/(row height)
When a CCD chip is being used as detector 141, the row height is the binned CCD row height.
It will be appreciated that integral multiples of nz may also be used for certain scanning techniques. Further, it will also be appreciated that various of the foregoing equations are susceptible of modification in accordance with system requirements, for example, or in accordance with individual limitations of system components. For example, operational characteristics of one or more stepper motors or servos implemented in stage or optics component motion control may dictate the values for nz or ny in some instances. Additionally or alternatively, some circumstances may require or benefit from dynamic computation or adjustment of nz and ny; one such circumstance may include, for example, acquisition of image data at a plurality of wavelengths during multi-spectral scans.
In that regard, motion control of various system components may be provided or accommodated by electronics 150 as set forth above with reference to
Turning now to the sequence illustrated in
The shading in the representation of the object to be imaged is provided to indicate the number of times a particular location of the object has been imaged by the image plane. The partially exploded view (i.e., the separation between the object and the image plane) is provided for clarity.
For example, when at scan position 1 in
It will be appreciated that at scan position 4, for example, and at other points during the scan sequence, the image plane may reverse direction; accordingly, z translation may be omitted at this point in the sequence. As indicated in
In the exemplary digital MAI scan of
In the foregoing embodiment, the equation nz=3ny−1 may be satisfied where the number of z increments in the scan sequence equals eleven (nz=11); that is, the twelve sequential scan positions represent eleven incremental movements in the z direction, starting from scan position 1.
As noted briefly above, integral multiples of nz may be employed in certain embodiments effectively to increase the z scan frequency. Alternative computations may also be used to calculate nz as a function of ny, additional factors, or some combination thereof in some embodiments. As with the analog MAI scan embodiments described above, digital MAI scans techniques may employ a variety of alternative z scan frequencies depending upon system requirements, component capabilities, multi-wavelength scan strategies, and so forth. Motion parameters such as lateral velocity and z scan frequency may be dynamically adjusted or selectively altered, for example, by control electronics 150 or similar programmable electronic components. In some embodiments, for example, electronics 150 may selectively adjust motion control parameters in accordance with feedback, data, or other instructions transmitted by image processor 142.
Upon examination of
Aspects of the present invention have been illustrated and described in detail with reference to particular embodiments by way of example only, and not by way of limitation. Those of skill in the art will appreciate that various modifications to the exemplary embodiments are within the scope and contemplation of the present disclosure. It is intended, therefore, that the present invention be limited only by the scope of the appended claims.
|Patente citada||Fecha de presentación||Fecha de publicación||Solicitante||Título|
|US4383327 *||1 Dic 1980||10 May 1983||University Of Utah||Radiographic systems employing multi-linear arrays of electronic radiation detectors|
|US4584704 *||1 Mar 1984||22 Abr 1986||Bran Ferren||Spatial imaging system|
|US4661986 *||18 May 1984||28 Abr 1987||Rca Corporation||Depth-of-focus imaging process method|
|US5231443 *||16 Dic 1991||27 Jul 1993||The Research Foundation Of State University Of New York||Automatic ranging and automatic focusing|
|US5889582 *||10 Mar 1997||30 Mar 1999||Virtek Vision Corporation||Image-directed active range finding system|
|US6201899 *||15 Ene 1999||13 Mar 2001||Sarnoff Corporation||Method and apparatus for extended depth of field imaging|
|US6320979 *||6 Oct 1998||20 Nov 2001||Canon Kabushiki Kaisha||Depth of field enhancement|
|US6418243 *||7 Mar 1997||9 Jul 2002||B. Ulf Skoglund||Apparatus and method for providing high fidelity reconstruction of an observed sample|
|US6522777 *||8 Jul 1999||18 Feb 2003||Ppt Vision, Inc.||Combined 3D- and 2D-scanning machine-vision system and method|
|US6549607 *||28 Abr 2000||15 Abr 2003||Wake Forest University||Method and system for creating task-dependent three-dimensional images|
|WO2003079664A2||13 Mar 2003||25 Sep 2003||Applied Precision, Llc||Multi-axis integration system and method|
|1||Agard et al., "Three-dimensional architecutre of a polytene nucleus", 1983, Nature, 302: 676-681.|
|2||Bracewell, The Fourier Transform and Its Applications, 1965, McGraw-Hill Publishing Company, pp. 24-27.|
|Clasificación de EE.UU.||356/601|
|Clasificación internacional||G06T1/00, G01B11/24|
|Clasificación cooperativa||H04N5/349, G02B21/367, G06T1/0007, H04N5/3743, H04N5/37206|
|Clasificación europea||H04N3/15D, H04N3/15H, G02B21/36V1, H04N5/374C, H04N5/372A, G06T1/00A|
|13 Jun 2003||AS||Assignment|
Owner name: APPLIED PRECISION, LLC, WASHINGTON
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:REESE, STEVEN A.;BROWN, CARL S.;GOODWIN, PAUL C.;REEL/FRAME:014181/0339
Effective date: 20030527
|4 Ago 2008||AS||Assignment|
Owner name: SILICON VALLEY BANK, CALIFORNIA
Free format text: SECURITY AGREEMENT;ASSIGNOR:APPLIED PRECISION, INC.;REEL/FRAME:021328/0695
Effective date: 20080429
Owner name: SILICON VALLEY BANK,CALIFORNIA
Free format text: SECURITY AGREEMENT;ASSIGNOR:APPLIED PRECISION, INC.;REEL/FRAME:021328/0695
Effective date: 20080429
|11 Sep 2008||AS||Assignment|
Owner name: APPLIED PRECISION, INC., WASHINGTON
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:APPLIED PRECISION, LLC;REEL/FRAME:021511/0562
Effective date: 20080429
|17 Mar 2011||FPAY||Fee payment|
Year of fee payment: 4
|3 Jun 2011||AS||Assignment|
Owner name: APPLIED PRECISION, INC., CALIFORNIA
Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:SILICON VALLEY BANK;REEL/FRAME:026398/0420
Effective date: 20110531
|3 Abr 2014||AS||Assignment|
Owner name: GE HEALTHCARE BIO-SCIENCES CORP., NEW JERSEY
Free format text: MERGER;ASSIGNOR:APPLIED PRECISION, INC.;REEL/FRAME:032595/0367
Effective date: 20130701
|16 Abr 2015||FPAY||Fee payment|
Year of fee payment: 8